The agent framework where every run is durable, replayable, and resumable by default.

A GPipe implementation in PyTorch

An I/O benchmark for deep Learning applications

Sayiir: simple, embeddable durable workflow engine in Rust with Node.js & Python bindings. Checkpoint-based recovery, no deterministic replay. A simplified alternative to Temporal, Restate, and Airflow.

CLI to the Cedana Service

Very-Low Overhead Checkpointing System

Extending DOLFINx with checkpointing functionality

Keras wrapper that autosaves what ModelCheckpoint cannot.

Execution runtime for intelligent agents with event-sourced, recoverable task orchestration.

Git-like branching, checkpointing, and comparison for AI agent execution paths. pip install agentgit

Weavegraph rust graph/agent/node

A platform for running sandboxed containers

An operating system for autonomous research — from literature to manuscript inside a governed, checkpointed loop.

Run Codex and Claude Code reliably in the background on macOS—with a native dashboard, persistent tmux sessions, local checkpoints, recovery, notifications, and closed-lid support.

Independent LangGraph C++ runtime for stateful AI agents and recoverable graph workflows: checkpointing, streaming, interrupts, tool calling, edge/local LLM apps. C++23 developer preview; not official LangGraph/LangChain.

[WIP] Debug TypeScript/JavaScript via TUI. Checkpoint functions, edit state, skip execution. Written in Rust 🦀

A Python package for checkpointing, saving, and loading objects.

Minimal PyTorch training framework — implement three methods, get a full training loop: composable phases, checkpointing, early stopping, metrics, and a live web dashboard.

A Python package for performing memory-intensive computations in parallel using chunks and checkpointing.

Polyglot AI agent orchestration core (Rust reference + TS/Go/.NET/Python bindings) — agents, workflows, tools, checkpointing, memory, HITL, cost. A security-first LangGraph; the engine behind smooth-operator and lom.smoo.ai.